HASI: Hardware-Accelerated Stochastic Inference, A Defense Against Adversarial Machine Learning Attacks

TL;DR

HASI introduces a hardware-accelerated stochastic inference method that detects adversarial inputs in DNNs by injecting noise during inference, achieving high detection rates with low performance overhead.

Contribution

The paper proposes a novel hardware-accelerated defense mechanism using stochastic inference to detect adversarial inputs efficiently.

Findings

Detection rate of 86% on VGG16

Detection rate of 93% on ResNet50

Overhead reduced to 1.58X-2X with co-designs

Abstract

Deep Neural Networks (DNNs) are employed in an increasing number of applications, some of which are safety critical. Unfortunately, DNNs are known to be vulnerable to so-called adversarial attacks that manipulate inputs to cause incorrect results that can be beneficial to an attacker or damaging to the victim. Multiple defenses have been proposed to increase the robustness of DNNs. In general, these defenses have high overhead, some require attack-specific re-training of the model or careful tuning to adapt to different attacks. This paper presents HASI, a hardware-accelerated defense that uses a process we call stochastic inference to detect adversarial inputs. We show that by carefully injecting noise into the model at inference time, we can differentiate adversarial inputs from benign ones. HASI uses the output distribution characteristics of noisy inference compared to a non-noisy reference to detect adversarial inputs. We show an adversarial detection rate of 86% when applied to VGG16 and 93% when applied to ResNet50, which exceeds the detection rate of the state of the art approaches, with a much lower overhead. We demonstrate two software/hardware-accelerated co-designs, which reduces the performance impact of stochastic inference to 1.58X-2X relative to the unprotected baseline, compared to 15X-20X overhead for a software-only GPU implementation.